✗ Failed on benchmark
2026
Replace magnitude pruning in a trained recurrent network with stochastic pruning probabilities computed from weight magnitudes and the covariance of neuron activities under injected noise. Connections whose endpoints fluctuate in a sign-compatible way receive higher retention probability, while retained weights are rescaled to preserve average recurrent strength. The method uses local weights and activity covariance, avoiding Hessian construction and expensive global saliency optimization.
Useful7/10
Difficulty4/10
Novelty7/10
✗ Failed on benchmark
2026
Construct optimizer variables as interconnected Hamiltonian subsystems: parameters store potential energy, momentum stores kinetic energy, and a skew coupling transfers energy between them without net creation. Positive-semidefinite resistance removes energy and provides an explicit damping knob, separating conservative exploration from dissipative convergence.
Useful7/10
Difficulty6/10
Novelty6/10
△ Mechanism confirmed, baseline not beaten
2026
Replace the standard diagonal or identity preconditioner used when solving an implicit neural layer with a coarse/fine Schur-complement preconditioner. The hidden state is decomposed into a low-dimensional coarse subspace and its orthogonal complement; the coarse interaction is solved accurately, while the fine block receives a damped approximate inverse. The method is especially suitable for deep equilibrium models, implicit MLPs, and Newton or quasi-Newton training of residual dynamics.
Useful7/10
Difficulty6/10
Novelty7/10
△ Mechanism confirmed, baseline not beaten
2026
Use the paper's asynchronous incremental aggregation pattern to train an orthogonal low-rank projection inside a neural network. Each worker refreshes only its local covariance-gradient cache when a minibatch arrives; the server aggregates cached ambient matrices and applies a polar retraction, so delayed workers do not require tangent-space transport or a global synchronization barrier. The resulting layer can support activation compression, online whitening, or a trainable low-rank bottleneck.
Useful7/10
Difficulty5/10
Novelty7/10
△ Mechanism confirmed, baseline not beaten
2026
Compress an existing dense neural-network weight matrix into a recursive butterfly operator using Gaussian sketches of complementary blocks. This is useful for deployment or distillation: the dense model provides an oracle for matrix-vector products, while the compressed model stores only recursive transfer bases and small cores. The generalized Nyström identity gives exact reconstruction for rank-k blocks and a principled approximation route for numerically low-rank blocks.
Useful7/10
Difficulty7/10
Novelty6/10
✓✓ Beats tuned baseline
2026
Replace a single global learning rate with mode-dependent rates determined by the static correlation structure of recent parameter updates or hidden-state updates. Correlated modes are treated as collective diffusive modes: their effective relaxation rate is reduced in proportion to their structure-factor amplitude, so the optimizer accelerates weakly correlated modes while damping collective slow modes. The method also supplies a diagnostic for when the Markovian approximation is invalid and…
Useful7/10
Difficulty6/10
Novelty7/10
✓✓ Beats tuned baseline
2026
Split a neural ODE or diffusion-model probability-flow ODE into a stiff known smoothing operator, a learned drift, and an optional local reaction term. Use super-time-stepping stages for the smoothing operator inside a single macrostep, while evaluating the learned drift only at selected coupling stages and treating the local reaction with diagonal or block-local implicit solves. This should allow substantially larger stable macrosteps when the known operator has a large negative spectral…
Useful7/10
Difficulty6/10
Novelty7/10
△ Mechanism confirmed, baseline not beaten
2026
Replace the exact matrix-polar normalization in Muon with the smoothed feedback \(h_\epsilon(M)=M(M^\top M+\epsilon I)^{-1/2}\). This retains singular-vector-aware updates and approximately unit-normalizes dominant spectral modes, but avoids unstable behavior when the momentum matrix is rank deficient or has tiny singular values.
Useful7/10
Difficulty5/10
Novelty4/10
✓✓ Beats tuned baseline
2026
Replace sequential upward message passing in a tree-structured neural module with rake–compress contraction of quadratic latent-state messages. Each node stores a quadratic value function and each edge stores a linear transition or coupling triple; leaf elimination and unary-node compression are implemented as batched Schur complements, followed by a reverse pass that reconstructs node latents and edge outputs. The layer is exact for Gaussian or quadratic latent models and remains…
Useful7/10
Difficulty6/10
Novelty7/10
△ Mechanism confirmed, baseline not beaten
2026
Use the relaxed QFT tensor-network topology as a trainable norm-preserving mixer inside a neural block, replacing a dense token-mixing matrix or an expensive global convolution. The network learns data-adapted global interactions while retaining structured O(N log^2 N) application and an exact cheap inverse, making it suitable for image tokens, long sequences, or reversible residual blocks.
Useful7/10
Difficulty6/10
Novelty6/10
✗ Failed on benchmark
2026
Replace unconstrained neural-network updates by updates projected toward directions supported by a recent, regularized gradient or feature subspace. This transfers PRPC's errors-in-variables correction: directions that are weakly identified by noisy or rank-deficient minibatches receive stronger shrinkage, preventing large updates caused by accidental correlations. The method is especially suitable for recurrent, world-model, and small-data fine-tuning problems where minibatch covariance is…
Useful7/10
Difficulty6/10
Novelty6/10
△ Mechanism confirmed, baseline not beaten
2026
Replace a single parameter iterate by two coupled replicas with unequal cross-couplings: replica 1 receives a force proportional to k_1(theta_1-theta_2), while replica 2 receives a force proportional to k_2(theta_2-theta_1), with k_1 not equal to k_2. The asymmetric coupling creates a controlled circulating component in the stochastic training dynamics, potentially helping escape flat saddles or correlated minibatch-noise traps without requiring an external periodic schedule. The coupling must…
Useful7/10
Difficulty5/10
Novelty7/10
△ Mechanism confirmed, baseline not beaten
2026
Add a controlled antisymmetric component to the local parameter update so optimization can circulate around ill-conditioned valleys instead of moving only along gradient directions. The symmetric component supplies dissipation, while the skew component produces the oscillatory non-reciprocal response predicted by the paper. Adapt the skew strength only while the estimated discrete-time dynamics remain stable.
Useful7/10
Difficulty6/10
Novelty7/10
✓✓ Beats tuned baseline
2026
Build an implicit or unrolled graph neural operator on a symmetric simplicial mesh, and perform every symmetry-compatible linear solve in a fixed representation-theoretic basis rather than the original edge/face basis. The same basis can be reused for Poisson, Maxwell, diffusion, and learned linear combinations of DEC operators, yielding parallel independent blocks and lower peak memory without changing the discretized solution.
Useful7/10
Difficulty6/10
Novelty6/10
△ Mechanism confirmed, baseline not beaten
2026
Replace explicit Runge-Kutta integration in a neural ODE or probability-flow ODE sampler with the anchored two-derivative method. Each stage uses both the neural vector field and its total time derivative, while the coupled implicit solve is designed so the accepted map has an L-stable Padé stability function. The method should allow larger steps on stiff trajectories without amplifying fast decaying modes.
Useful7/10
Difficulty7/10
Novelty7/10
✗ Failed on benchmark
2026
Replace ordinary gradient descent in a chosen approximately linear parameter block with gradient descent plus a controlled negative quadratic penalty, and stop before the unstable directions explode. The finite-time spectral filter can amplify well-supported directions while retaining shrinkage or limited exposure on weak directions, which is unavailable to a stable negative-ridge endpoint.
Useful7/10
Difficulty5/10
Novelty7/10
✓✓ Beats tuned baseline
2026
Train a mixture-of-experts router by solving its regularized nonnegative simplex least-squares subproblem with a matrix-free active-set conjugate-gradient method instead of projected gradient or Adam. The router coefficients remain exactly nonnegative and sum to one, while CG rapidly solves each free-set quadratic and the active-set pivots identify sparse expert assignments.
Useful7/10
Difficulty5/10
Novelty7/10
△ Mechanism confirmed, baseline not beaten
2026
Replace a portion of quadratic key-value attention or an external episodic table with a per-sample matrix fast memory updated by rank-one delta corrections. The memory directly learns a linear key-to-value map and can be carried across sequence segments, providing cheap online adaptation with constant state size per head.
Useful7/10
Difficulty5/10
Novelty5/10
△ Mechanism confirmed, baseline not beaten
2026
Build an SPD classifier and residual head directly from Cholesky factors, using lower-triangular differences and matrix-power terms instead of generic eigendecomposition-based logarithm operators. This retains covariance geometry while making positive-definiteness automatic and backpropagation more numerically stable for minibatch training.
Useful7/10
Difficulty6/10
Novelty7/10
✗ Failed on benchmark
2026
Construct overlapping attention windows from a block design instead of using one dense sequence-by-sequence attention matrix. Every token appears in exactly $r$ windows and every token pair co-occurs in exactly $\lambda$ windows, giving uniform coverage and avoiding the uneven connectivity of arbitrary sparse masks.
Useful7/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
Split hidden dynamics into relaxation bands when the Jacobian spectrum has a gap, evolve each band with its own timescale, and retain an explicit cross-band exchange term. This yields a principled dual-timescale RNN or SSM rather than choosing fast and slow branches heuristically.
Useful7/10
Difficulty6/10
Novelty6/10
✗ Failed on benchmark
2026
Replace a stationary optimizer by a periodic two- or multi-phase schedule, such as alternating large and small learning rates, SGD and momentum, or gradients from different loss components. Stability is assessed over the complete period using the product of phase-wise linearized update maps, allowing a phase that is individually expansive to be safely combined with a contracting phase.
Useful7/10
Difficulty5/10
Novelty7/10
△ Mechanism confirmed, baseline not beaten
2026
Grow a neural network by appending a trainable block together with an analytically initialized inverse block, so the newly added depth is exactly the identity at insertion time. After insertion, untie and optimize the two blocks independently; this preserves the current function while providing additional trainable degrees of freedom. For architectures with one expensive mixing operation followed by cheap channelwise blocks, the same construction can increase depth without repeatedly paying for…
Useful7/10
Difficulty5/10
Novelty5/10
✗ Failed on benchmark
2026
Compress a transformer KV cache by selecting actual past tokens whose key or hidden-state columns form a stable basis for all cached tokens. Instead of retaining tokens with the largest attention scores or leverage scores independently, compute rank-revealing pivoting of the leading right-singular-vector matrix and retain its pivot columns, then evaluate attention using the representatives plus an optional low-cost residual correction.
Useful7/10
Difficulty5/10
Novelty6/10